Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
×

Comment Re: Failsafe? (Score 1) 468

When the instruments are gone (somehow), the pilot would certainly rather have windows than not. It provides a very good chance of survivability where no windows means practically no chance.

I'm not getting on a plane that has no backup plan.

so you won't get on an Airbus already?

Seriously, Airbus designs their planes so that the computers override the pilots. If the pilot wants to do something that the computer doesn't think they should do, then they have to enter override codes to be allowed to do so - which means, in an emergency, that the pilots have to spend precious time overriding the computer to resolve the emergency.

So honestly this is no surprise from Airbus; just a natural evolution of their already computer controlled systems with people as a secondary system.

Comment Re:Cost (Score 1) 143

Slightly different beasts I think. R is a really impressive analysis tool. Python is a scripting language. The latter is quite a bit more versatile, but ... probably isn't the right tool to solve the problem outlined in the OP.

However, his question was related to coming from SAS. SAS Scripting is not a general language eithers; it very much like using GNU Octave, Mathamatica, Matlab, and R - able to do some general things (open/read/write/close files) but is generally very data set oriented. So R is very much a suitable replacement.

Comment Re:Belief vs Experience (Score 1) 143

The cost of training them to use R will be signifantly cheaper than what you are spending on the SAS licenses And yes, while I have not used R myself, I would certainly recommend it over Python for this use case

So not having used R yourself, why do you believe it is the better and cheaper solution?

B/c I don't do much in data modeling and working with that kind of data. The SAS scripts I wrote were over 10 years ago (2002). R isn't that old. I had one time which it might have (last summer), but then forgot about it. But for me that's 1 time in >10 years. If I needed to get into doing the stuff I did with SAS again, then yes I'd be looking at using R; but that's unlikely for me.

And yes, I have seen others use it so I know how much easier it would be for me to get into using R than updating myself on SAS and getting back into that.

Comment Re:Scientific research never got anyone anything (Score 1) 225

It's not even that. The military is getting their budget cut the same as every other government agency. A more accurate statement would be:

"Still, I guess there are budget hawks who need to get re-elected, so something had to give."

Well that is not fair, the military's budget is so colossal that they should be cut at a much higher rate than everything else.

The militar budget is a very small fraction of the entire budget - something like 4%.

If you want to talk about colossal budgets then look at entitlements - Healthcare, Medicare, Medicaid, Welfare, etc - which compromise over 50%.

Comment Re:why? (Score 1) 346

The real question is: should the court order such an action, and under what conditions? Analogy alert: GS mistakenly sends me a letter by physical mail, then asks the post office (or asks a judge to order the post office) to send a mailman round, break into my house, and retrieve the letter. That clearly won't happen; worst case is that the judge would order me to surrender the letter. In case of email, is Google (under their terms & conditions and the letter of the law) allowed to "break into" my mailbox and remove the offending letter? And should they be?

You analogy would be better if the mail had been left in the mailbox, which is regulated by the USPS and Federal Law, and which the postman has rights to access.

So it would be more like:

GS sends you a letter by mistake. They get a court order to order the USPS to remove it from the mailbox that they put in it, which happens to be yours. The postman then looks at the contents of the mailbox, verifies it is still there, and then removes it, sending it back to the sender or as otherwise directed by the courts. If, however, you checked your mail and took the mail out and into your house, then there is nothing for the USPS to do - it is no longer in the mailbox. If, however, you keep all your mail in your mailbox then the USPS would be within their ability to remove it from the mailbox.

So, to keep all your e-mail on Google's servers (or any ISPs servers) opens up the opportunity for this to happen. To keep the opportunity from happening then you need to download your e-mail from anyone else's servers and store it locally, deleting it from the providers servers after you have your local copy.

Alternatively, you can rent your own server and host your own e-mail server; but then you get into another situation in which the service you are renting from may be required by the courts to turn over the server to the courts, or get shutdown and you've now lost everything except what you had backups for. And yes, that has happened where the FBI shutdown a hosting provider and took all the servers in order to get to one of their clients; everyone else was screwed for a while.

Comment Re:If they approve allowing calls on planes... (Score 1) 128

That will be the last time I fly commercial. The LAST thing I want to do is be couped up in an aluminum can for 1+ hours listening to half of other people's mindless drivel conversations on their phones. It's already bad enough the second the plane hits the runway on landing everyone pulls out their phones to call people. And they don't just have the "ok we just landed I'll meet you out front in 20 minutes" short talk. - No it turns into long drawn out annoying conversations hat CERTAINLY can wait until they are off the plane to have.

In-flight phones in the back of the seat have been available for years. Were in-flight calls a problem for other passengers, you'd think we would have realized it by now.

Those phones were also largely ignored because of the expense. Of course, they didn't necessarily communicate to the cell towers like your phone does either - they probably went through the plane's standard communications mechanisms (f.e sat-com, etc.) and then got routed out to a phone system on the ground.

Comment Re:If they approve allowing calls on planes... (Score 1) 128

That will be the last time I fly commercial. The LAST thing I want to do is be couped up in an aluminum can for 1+ hours listening to half of other people's mindless drivel conversations on their phones. It's already bad enough the second the plane hits the runway on landing everyone pulls out their phones to call people. And they don't just have the "ok we just landed I'll meet you out front in 20 minutes" short talk. - No it turns into long drawn out annoying conversations hat CERTAINLY can wait until they are off the plane to have.

TSA is nearly enough for me to do that already.

Comment Cost (Score 4, Informative) 143

The cost of training them to use R will be signifantly cheaper than what you are spending on the SAS licenses, which (last I knew) was a yearly purchase for each user.

And yes, while I have not used R myself, I would certainly recommend it over Python for this use case as it is very dedicated to doing the kinds of things that SAS is good at in a very efficient, friendly manner. I've seen a number of people use it to do some very neat statistical analysis, and their stuff was a lot simpler than the SAS scripts that I use to write years back.

Comment Re:sound and sides (Score 1) 579

I think the gp is thinking of large loaded trucks that have to go through a few gears just to make it through the intersection and might leave the intersection at 10-15 MPH. These trucks also usually try to avoid actually stopping when approaching the intersection, rather slowing down to a couple of MPH

Yes, and they will often blow the horn signaling that they are going to run the red light because they may not be able to stop in time either.

That said, even when driving a car, I find the counters very useful to help determine from the cross direction (e.g the direction already with the green light) when the light may change when I'm a good distance off. Sometimes they're very good predictors of the light about to go yellow so I can slow down instead.

Making them unreadable will have a great unintended consequence of pedestrians - especially smaller folks and children - unable to read them at all - you know, the very people they were put there to help protect.

Comment Re:Somebody has to do it (Score 1) 178

This hole is easily found and defeated, provided you have two independent compilers. You don't actually have to trust either, only that they aren't jiggered in the same way.

Say you suspect compiler A. Take its source (A') and compile it with compilers A and B. Let's call the results A(A') and B(A'). Since A and B doubtless do different things, there's likely to be a whole lot of differences, so you won't be able to tell if there's a backdoor in A, although if there is it will be in A(A') and not B(A').

Thing is, since A' is the source for a compiler, A(A') and B(A') are compilers, and since both A and B compile the same language they should do the same thing, agreeing with the language semantics of A' - assuming, of course, that there's nothing underhanded going on. Therefore, we can use these newly compiled compilers to compile A', getting (A(A'))A' and (B(A'))A'. These should be identical, since they were compiled by compilers that ostensibly do the same thing. If they're not, you've found a problem, and you can use (B(A'))A' instead of A in the future.

IIRC per bulding GCC, I believe GCC does some of that as part of its build process - it builds a version of itself to build itself so the executable you finally get is built by itself not the system compiler. It's been a while, but I believe its doing that even if you are not doing cross-platform builds.

Sure you don't have a second compiler to compare against, but it's a pretty good guarantee that the compiler is what the code said it is.

Comment Re:They are not a charity (Score 2) 228

My read of this is that they applied as a charity, but the IRS's definition of a charity requires that you be serving a distinct, disadvantaged group of people. A quick look at the software that Yorba produces (http://yorba.org), does not lead me to believe that their software would particularly benefit any specific disadvantaged groups more than other people.

So by the rules that the IRS is working on, it does appear that they do not qualify as a charity. And to be honest, this is a correct definition, they are not running a charity. Now there is a valid question about whether there should be a method for them to run a non-profit without being taxes, but they are not a charity.

There are many kinds of Charitable organizations. But 501(c)3 does not necessarily mean a Charity as you describe, though it does allow you to take donations. Most of the 501(c) organizations are pretty specific in what they may serve; 501(c)3 is the exception in that it is a lot more general.
The Wikipedia Article on 501(c) organizations is actually pretty good. Of course, you can also go directly to the IRS information too, but I find the Wikipedia article to be easier to read.

Comment Re:WTF (Score 1) 495

And this is why we need Namecoin and other decentralized DNS solutions to take such matters out of the hands of the lawmakers.

http://namecoin.info/

It won't take it out of the hands of lawmakers. It may make it harder to regulate, but they'll find a way - even if it is requiring all ISPs and Back bone providers to block it.

It's naive to think that just because its decentralized it is beyond the reach of government (or even corporations for that matter).

Slashdot Top Deals

The rule on staying alive as a program manager is to give 'em a number or give 'em a date, but never give 'em both at once.

Working...